Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881359
T. Addabbo, Federico Carli, A. Fort, Federico Micheletti, E. Panzardi, V. Vignoli
The use of contactless sensing systems represents a measurement technique of great potentiality in the development of engineering systems that have to face specific challenges related to hazardous or critical measurement environment. In this paper a telemetric sensing system for QCM sensors is proposed. The system is based on air coupled antennas, a timed excitation, and an optimized readout circuit, properly designed to avoid disruptive loading of the resonant system. The measurement strategy involves the excitation of the quartz with a sine burst and the acquisition of the transient response after the end of the excitation phase. The excitation system implements a strategy for the optimal excitation frequency search, which allows for the estimation of the resonance frequency of the QCM in the different operating conditions also in the presence of large mechanical loads as it occurs in in-liquid measurements. In this way the resonant electromechanical system is forced to operate close to its optimum working point with the maximum signal-to-noise ratio and preserve the typical high performance of QCM based monitoring systems in terms of sensitivity and frequency stability.
{"title":"Telemetric QCM-D based sensing system with adaptive excitation frequency","authors":"T. Addabbo, Federico Carli, A. Fort, Federico Micheletti, E. Panzardi, V. Vignoli","doi":"10.1109/SAS54819.2022.9881359","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881359","url":null,"abstract":"The use of contactless sensing systems represents a measurement technique of great potentiality in the development of engineering systems that have to face specific challenges related to hazardous or critical measurement environment. In this paper a telemetric sensing system for QCM sensors is proposed. The system is based on air coupled antennas, a timed excitation, and an optimized readout circuit, properly designed to avoid disruptive loading of the resonant system. The measurement strategy involves the excitation of the quartz with a sine burst and the acquisition of the transient response after the end of the excitation phase. The excitation system implements a strategy for the optimal excitation frequency search, which allows for the estimation of the resonance frequency of the QCM in the different operating conditions also in the presence of large mechanical loads as it occurs in in-liquid measurements. In this way the resonant electromechanical system is forced to operate close to its optimum working point with the maximum signal-to-noise ratio and preserve the typical high performance of QCM based monitoring systems in terms of sensitivity and frequency stability.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"2006 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125841531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881349
T. Polonelli, Andrea Bentivogli, Guido Comai, M. Magno
Unexpected equipment failure is expensive and potentially hazardous for workers and users. Periodic inspections and maintenance at predefined intervals aim to limit unplanned production downtime, costly replacement of parts and safety concerns. On the other side, predictive maintenance techniques can monitor equipment as it operates, anticipating deterioration and incoming breakages, enabling just-in-time services at reduced operational costs. This paper presents a deploy and forget predictive maintenance sensor node designed explicitly for industrial electric motors. It is targeted for AC mono and three-phase asynchronous motors and generators, measuring vibrations, environmental noise, temperature, and the external magnetic field. The proposed sensor achieves self-sustainability by exploiting a 4x4 cm thermal source for 72 s with a ∆T of 20 °C, and it features short-long wireless data transfer respectively over WiFi and the cellular NB-IoT network. We tested the prototype on different electric motors, form 4 kW to 110 kW, reporting here its capability to detect anomalies using a vibration spectral analysis.
{"title":"Self-sustainable IoT Wireless Sensor Node for Predictive Maintenance on Electric Motors","authors":"T. Polonelli, Andrea Bentivogli, Guido Comai, M. Magno","doi":"10.1109/SAS54819.2022.9881349","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881349","url":null,"abstract":"Unexpected equipment failure is expensive and potentially hazardous for workers and users. Periodic inspections and maintenance at predefined intervals aim to limit unplanned production downtime, costly replacement of parts and safety concerns. On the other side, predictive maintenance techniques can monitor equipment as it operates, anticipating deterioration and incoming breakages, enabling just-in-time services at reduced operational costs. This paper presents a deploy and forget predictive maintenance sensor node designed explicitly for industrial electric motors. It is targeted for AC mono and three-phase asynchronous motors and generators, measuring vibrations, environmental noise, temperature, and the external magnetic field. The proposed sensor achieves self-sustainability by exploiting a 4x4 cm thermal source for 72 s with a ∆T of 20 °C, and it features short-long wireless data transfer respectively over WiFi and the cellular NB-IoT network. We tested the prototype on different electric motors, form 4 kW to 110 kW, reporting here its capability to detect anomalies using a vibration spectral analysis.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128467858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881360
G. Bandini, M. Marracci, G. Caposciutti, B. Tellini
In this paper we compare different magnetic field sensor technologies that can potentially be used to characterize current distribution in massive conductors. The approach is based on the application of the current barycenter method, introduced by the authors in a previous work, to obtain information about the current distribution in massive conductors fed by pulsed current through the use of an array of search loops placed around the conductor. After briefly recalling the method, this paper analyzes the possible use of different magnetic field sensor technologies (search loops, vector and scalar magnetometers) to compare their performance in the use of the current barycenter reconstruction method. The introduced model error as a function of sensor type, conductor cross-sectional shape, and mutual position between sensors and conductor are analyzed and discussed throughout the paper.
{"title":"Comparison of Magnetic Field Sensors for Current Distribution Reconstruction through Barycenter Filament Model","authors":"G. Bandini, M. Marracci, G. Caposciutti, B. Tellini","doi":"10.1109/SAS54819.2022.9881360","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881360","url":null,"abstract":"In this paper we compare different magnetic field sensor technologies that can potentially be used to characterize current distribution in massive conductors. The approach is based on the application of the current barycenter method, introduced by the authors in a previous work, to obtain information about the current distribution in massive conductors fed by pulsed current through the use of an array of search loops placed around the conductor. After briefly recalling the method, this paper analyzes the possible use of different magnetic field sensor technologies (search loops, vector and scalar magnetometers) to compare their performance in the use of the current barycenter reconstruction method. The introduced model error as a function of sensor type, conductor cross-sectional shape, and mutual position between sensors and conductor are analyzed and discussed throughout the paper.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130548849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881377
O. P. Maurya, P. Sumathi
The flexible capacitive sensors are employed for micro-droplet detection in drug delivery systems. The structures of capacitive sensors such as interdigitated electrode and parallel plate types possess certain geometrical properties which are greatly influencing the accuracy, sensitivity, and stability. These properties are improved in semi-cylindrical type, whereas it is nullified in cross-capacitive sensor. In this paper, the semi-cylindrical and cross-capacitive sensor designs are analyzed for the suitability of micro-droplet detection. The proposed simulation studies include the change of capacitance due to variation in liquid droplet size and free-flying liquid droplet position sweep, electric field distribution between sensing and working electrodes. Moreover, the guard electrodes and metal shielding effects are analyzed for further improvement in sensor performance by eliminating the effects of stray capacitance. The COMSOL Multiphysics based simulation studies reveal the suitable sensitivity and change in capacitance for the sensor design.
{"title":"Design Considerations of Capacitive Sensors for Micro-Droplet Detection","authors":"O. P. Maurya, P. Sumathi","doi":"10.1109/SAS54819.2022.9881377","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881377","url":null,"abstract":"The flexible capacitive sensors are employed for micro-droplet detection in drug delivery systems. The structures of capacitive sensors such as interdigitated electrode and parallel plate types possess certain geometrical properties which are greatly influencing the accuracy, sensitivity, and stability. These properties are improved in semi-cylindrical type, whereas it is nullified in cross-capacitive sensor. In this paper, the semi-cylindrical and cross-capacitive sensor designs are analyzed for the suitability of micro-droplet detection. The proposed simulation studies include the change of capacitance due to variation in liquid droplet size and free-flying liquid droplet position sweep, electric field distribution between sensing and working electrodes. Moreover, the guard electrodes and metal shielding effects are analyzed for further improvement in sensor performance by eliminating the effects of stray capacitance. The COMSOL Multiphysics based simulation studies reveal the suitable sensitivity and change in capacitance for the sensor design.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"514 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123249664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881338
E. Sisinni, A. Depari, P. Bellagente, P. Ferrari, A. Flammini, M. Pasetti, S. Rinaldi
As constantly stated by the World Health Organization, physical activity is extremely important for a healthy aging, but how exercises are made is as important as how much activity is made. A large variety of wearable devices capable of sensing people movement is available on the market. Automatic detection and classification of fitness activity is also possible, leveraging artificial intelligence (AI) algorithms. In this paper, some ideas on the impact of specific input features on AI model performance for fitness exercise recognition is reported and discussed. Then, a general classification of input features is proposed. Using a pre-recorded dataset composed of 9 exercise repetition sets performed by 7 volunteers, a LSTM network have been trained and validated using the Leave One Out Cross Validation approach. Finally, the same network has been re-trained several times, varying the input parameters. Differences in classification results due to such parameters have been evaluated through the precision, recall and accuracy metrics. In particular, the precision is between 97.8% and 63.8%, whereas recall is between 98.5% and 42.3%, in line with results in literature.
正如世界卫生组织(World Health Organization)不断强调的那样，体育锻炼对于健康的老龄化极其重要，但如何锻炼与运动量同样重要。市场上有各种各样能够感知人们运动的可穿戴设备。利用人工智能(AI)算法，健身活动的自动检测和分类也是可能的。本文报道并讨论了特定输入特征对健身运动识别AI模型性能影响的一些想法。然后，提出了输入特征的一般分类方法。使用由7名志愿者执行的9个练习重复集组成的预记录数据集，使用Leave One Out交叉验证方法对LSTM网络进行了训练和验证。最后，对同一个网络进行多次重新训练，改变输入参数。由于这些参数导致的分类结果差异已经通过精密度、召回率和准确度指标进行了评估。其中，准确率在97.8% ~ 63.8%之间，召回率在98.5% ~ 42.3%之间，与文献结果一致。
{"title":"On feature selection in automatic detection of fitness exercises using LSTM models","authors":"E. Sisinni, A. Depari, P. Bellagente, P. Ferrari, A. Flammini, M. Pasetti, S. Rinaldi","doi":"10.1109/SAS54819.2022.9881338","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881338","url":null,"abstract":"As constantly stated by the World Health Organization, physical activity is extremely important for a healthy aging, but how exercises are made is as important as how much activity is made. A large variety of wearable devices capable of sensing people movement is available on the market. Automatic detection and classification of fitness activity is also possible, leveraging artificial intelligence (AI) algorithms. In this paper, some ideas on the impact of specific input features on AI model performance for fitness exercise recognition is reported and discussed. Then, a general classification of input features is proposed. Using a pre-recorded dataset composed of 9 exercise repetition sets performed by 7 volunteers, a LSTM network have been trained and validated using the Leave One Out Cross Validation approach. Finally, the same network has been re-trained several times, varying the input parameters. Differences in classification results due to such parameters have been evaluated through the precision, recall and accuracy metrics. In particular, the precision is between 97.8% and 63.8%, whereas recall is between 98.5% and 42.3%, in line with results in literature.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124870049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881381
F. Paissan, V. Kumaravel, Elisabetta Farella
Electroencephalogram (EEG) signals recorded from the scalp are often affected by artifacts. Most existing artifact detection methods rely on multi-channel statistics such as inter-channel correlation. Recently, there has been a growing interest in realizing single-channel EEG systems to promote everyday use, demanding novel artifacts detection techniques. This paper presents validation results for single-channel artifacts detection in raw EEG signals using four neural architectures: a one-dimensional CNN (1D-CNN) - proposed by us, EEGNet, SincNet and EEGDenoiseNet. We used semi-synthetic EEG data corrupted with Ocular (EOG) and Myo-graphic (EMG) noise components to validate the approaches. Precisely, we contaminated the randomly chosen EEG channels with EOG and EMG artifacts in a controlled manner using a predefined Signal-to-Noise Ratio (SNR) such that the ground truth is known. We validated these models both in terms of classification performance and the interpretability of the learned features. Of the four models, 1D-CNN, EEGNet, and SincNet achieved a comparable classification accuracy (around 99%) and EEGDenoiseNet achieved as low as 64%. Analysing the learned filters for interpretability, we found both 1D-CNN and EEGNet clearly separates EOG (Delta and Theta) and EMG (Gamma) frequency bands from EEG. Instead, SincNet prioritized to learn EEG-specific features (Alpha and Beta) rather than artifact-related information still achieiving the comparable performance as the other two models. EEGDenoiseNet with kernel width of 3 was excluded from this evaluation as it is practically infeasible to perform FFT analysis. Finally, we also computed the number of training parameters for each model to evaluate which among them would be suitable for a resource-constrained wearable device and we found that 1D-CNN and SincNet are the most parameter-efficient, although not by a large margin.
{"title":"Interpretable CNN for Single-Channel Artifacts Detection in Raw EEG Signals","authors":"F. Paissan, V. Kumaravel, Elisabetta Farella","doi":"10.1109/SAS54819.2022.9881381","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881381","url":null,"abstract":"Electroencephalogram (EEG) signals recorded from the scalp are often affected by artifacts. Most existing artifact detection methods rely on multi-channel statistics such as inter-channel correlation. Recently, there has been a growing interest in realizing single-channel EEG systems to promote everyday use, demanding novel artifacts detection techniques. This paper presents validation results for single-channel artifacts detection in raw EEG signals using four neural architectures: a one-dimensional CNN (1D-CNN) - proposed by us, EEGNet, SincNet and EEGDenoiseNet. We used semi-synthetic EEG data corrupted with Ocular (EOG) and Myo-graphic (EMG) noise components to validate the approaches. Precisely, we contaminated the randomly chosen EEG channels with EOG and EMG artifacts in a controlled manner using a predefined Signal-to-Noise Ratio (SNR) such that the ground truth is known. We validated these models both in terms of classification performance and the interpretability of the learned features. Of the four models, 1D-CNN, EEGNet, and SincNet achieved a comparable classification accuracy (around 99%) and EEGDenoiseNet achieved as low as 64%. Analysing the learned filters for interpretability, we found both 1D-CNN and EEGNet clearly separates EOG (Delta and Theta) and EMG (Gamma) frequency bands from EEG. Instead, SincNet prioritized to learn EEG-specific features (Alpha and Beta) rather than artifact-related information still achieiving the comparable performance as the other two models. EEGDenoiseNet with kernel width of 3 was excluded from this evaluation as it is practically infeasible to perform FFT analysis. Finally, we also computed the number of training parameters for each model to evaluate which among them would be suitable for a resource-constrained wearable device and we found that 1D-CNN and SincNet are the most parameter-efficient, although not by a large margin.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"1999 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127485574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881254
S. S. Gilakjani, Hussein Al Osman
To determine the quality of experience for users of technological devices, we must consider the human influential factors, which encompass the emotional state. Hence, we propose a model to estimate user emotions from Electroencephalogram (EEG) signals. The model is a cascade of deep learning networks consisting of a pre-trained convolutional neural network which extracts spatial relations and residual block(s) of recurrent neural network which learn the temporal relations of multi-channel EEG signals and uses shortcuts across the neural layers for a more effective training of the deep network. We adopted the DEAP dataset to train and evaluate our model. To confirm that the proposed work is user-independent, we ensure that the data in the test set corresponds to subjects that are not included in the training set. We explored several input sets to determine the one that performs best on the DEAP dataset. We implemented existing popular state-of-the-art methods and compared with the proposed model. The results indicate that the proposed model consistently outperforms the previous models achieving 0.61 and 0.63 accuracy on the validation and 0.65 and 0.68 accuracy on the test dataset for valence and arousal respectively.
{"title":"Emotion Classification from Electroencephalogram Signals Using a Cascade of Convolutional and Block-Based Residual Recurrent Neural Networks","authors":"S. S. Gilakjani, Hussein Al Osman","doi":"10.1109/SAS54819.2022.9881254","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881254","url":null,"abstract":"To determine the quality of experience for users of technological devices, we must consider the human influential factors, which encompass the emotional state. Hence, we propose a model to estimate user emotions from Electroencephalogram (EEG) signals. The model is a cascade of deep learning networks consisting of a pre-trained convolutional neural network which extracts spatial relations and residual block(s) of recurrent neural network which learn the temporal relations of multi-channel EEG signals and uses shortcuts across the neural layers for a more effective training of the deep network. We adopted the DEAP dataset to train and evaluate our model. To confirm that the proposed work is user-independent, we ensure that the data in the test set corresponds to subjects that are not included in the training set. We explored several input sets to determine the one that performs best on the DEAP dataset. We implemented existing popular state-of-the-art methods and compared with the proposed model. The results indicate that the proposed model consistently outperforms the previous models achieving 0.61 and 0.63 accuracy on the validation and 0.65 and 0.68 accuracy on the test dataset for valence and arousal respectively.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"214 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133511432","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881380
Carla Amatetti, T. Polonelli, Enea Masina, Charles Moatti, D. Mikhaylov, D. Amato, A. Vanelli-Coralli, M. Magno, L. Benini
Internet of Things devices and smart sensors have become increasingly more pervasive in railway transportation applications, where they have the potential to significantly improve reliability, capacity, safety, and to reduce costs. In the ‘smart rail’ concept a key enabler is the ability to accurately localize trains with centimeter precision. This can be achieved using a combination of a high-precision GNSS-based module capable of achieving sub-meter accuracy and emerging radio and sensor technologies. This paper proposes a train tracking sensor node for in-field assessments fusing the absolute localization data from the GNSS and from local reference systems, such as Real Time Kinematics (RTK) with Inertial Measurement Unit (IMU) and Dead Reckoning (DRK). A complete wireless sensor node has been designed and evaluated in the field for functionality and power consumption. Within the sensor node, two different GNSS modules have been tested, with and without RTK and DRK, under different GNSS coverage conditions in various static and dynamic scenarios. We demonstrate that centimeter accuracy is achievable, with an accuracy of 2 ± 1 cm under static conditions and perfect satellite visibility, 4 ± 18 cm and 17 ± 40 cm under dynamic conditions in perfect and poor coverage conditions, respectively.
{"title":"Towards the Future Generation of Railway Localization and Signaling Exploiting sub-meter RTK GNSS","authors":"Carla Amatetti, T. Polonelli, Enea Masina, Charles Moatti, D. Mikhaylov, D. Amato, A. Vanelli-Coralli, M. Magno, L. Benini","doi":"10.1109/SAS54819.2022.9881380","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881380","url":null,"abstract":"Internet of Things devices and smart sensors have become increasingly more pervasive in railway transportation applications, where they have the potential to significantly improve reliability, capacity, safety, and to reduce costs. In the ‘smart rail’ concept a key enabler is the ability to accurately localize trains with centimeter precision. This can be achieved using a combination of a high-precision GNSS-based module capable of achieving sub-meter accuracy and emerging radio and sensor technologies. This paper proposes a train tracking sensor node for in-field assessments fusing the absolute localization data from the GNSS and from local reference systems, such as Real Time Kinematics (RTK) with Inertial Measurement Unit (IMU) and Dead Reckoning (DRK). A complete wireless sensor node has been designed and evaluated in the field for functionality and power consumption. Within the sensor node, two different GNSS modules have been tested, with and without RTK and DRK, under different GNSS coverage conditions in various static and dynamic scenarios. We demonstrate that centimeter accuracy is achievable, with an accuracy of 2 ± 1 cm under static conditions and perfect satellite visibility, 4 ± 18 cm and 17 ± 40 cm under dynamic conditions in perfect and poor coverage conditions, respectively.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132175246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-08-01DOI: 10.1109/SAS54819.2022.9881352
R. Yan, H. Viumdal, K. Fjalestad, S. Mylvaganam
in oil and gas industries. Accurately identifying flow types and estimating flow velocities of the individual phases are crucial for different purposes, such as observing the process status and providing inputs to control systems. This paper presents a solution for identifying flow contents and estimating flow rates in single-phase or each phase in multiphase flows by using pressure measurements and pipe vibrations caused by the flows. The necessary experiments were performed using the multiphase flow rig with three-inch diameter pipelines transporting natural gas, water, and crude oil in a closed loop with a separator tank as source and sink. A series of tree-based ensemble machine learning models have been developed and tested with the data collected from accelerometers, differential pressure transmitters, and upstream- and downstream pressure transmitters. With these inputs, the developed models can identify volume ratios of individual phases (such as water cut) and can estimate the flow velocity of each phase in the flow loop, including the open/close status of the choke valve. After describing briefly, the P&ID diagram of the multiphase flow rig, the paper focuses on exploratory data analysis of the data from three accelerometers and three pressure sensors using three submodels cascaded to perform ensemble learning.
{"title":"Ensemble learning in the estimation of flow types and velocities of individual phases in multiphase flow using non-intrusive accelerometers' and process pressure data","authors":"R. Yan, H. Viumdal, K. Fjalestad, S. Mylvaganam","doi":"10.1109/SAS54819.2022.9881352","DOIUrl":"https://doi.org/10.1109/SAS54819.2022.9881352","url":null,"abstract":"in oil and gas industries. Accurately identifying flow types and estimating flow velocities of the individual phases are crucial for different purposes, such as observing the process status and providing inputs to control systems. This paper presents a solution for identifying flow contents and estimating flow rates in single-phase or each phase in multiphase flows by using pressure measurements and pipe vibrations caused by the flows. The necessary experiments were performed using the multiphase flow rig with three-inch diameter pipelines transporting natural gas, water, and crude oil in a closed loop with a separator tank as source and sink. A series of tree-based ensemble machine learning models have been developed and tested with the data collected from accelerometers, differential pressure transmitters, and upstream- and downstream pressure transmitters. With these inputs, the developed models can identify volume ratios of individual phases (such as water cut) and can estimate the flow velocity of each phase in the flow loop, including the open/close status of the choke valve. After describing briefly, the P&ID diagram of the multiphase flow rig, the paper focuses on exploratory data analysis of the data from three accelerometers and three pressure sensors using three submodels cascaded to perform ensemble learning.","PeriodicalId":129732,"journal":{"name":"2022 IEEE Sensors Applications Symposium (SAS)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128973832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：